Autism spectrum disorders (ASD) are characterized by social impairments and are one of the most devastating childhood disorders, but remain poorly understood. Few robust biomarkers of ASD have been identified, hindering the understanding of basic biology; nor are there any pharmacotherapies that treat the social deficits of ASD. Progress has been impeded, in part, by the difficulty of obtaining relevant tissue samples from patients and their matched neurotypical controls. In mouse models, while tissue is available, there is not infrequently discordance between complex human behavior and laboratory-based mouse behavior, even with shared genetic etiologies. These two limitations underscore the tremendous value in developing an animal model of social impairments with more reliable behavioral and biological correlates to the human disease. Rhesus monkeys are an ideal model organism. Like humans, they are highly social, and both species display stable and pronounced individual differences in social functioning. At the behavioral extremes, low-sociable compared to high-sociable male rhesus monkeys initiate fewer affiliative interactions and display more inappropriate social behavior, suggesting both lower social motivation and poorer social skills. Naturally occurring low sociability in male rhesus monkeys therefore presents an exceptional opportunity to study the biology of social impairments. This research leverages a novel statistical classification model which predicts, from infant behavior, subjects that are subsequently found to be low-sociable or high-sociable. In this grant, we propose to collect quantitative social behavior data in a larger validation cohort of 1-5 year old male monkeys to confirm the discriminant power of this statistical model. If successful, this statistica model will serve as a robust high- throughput screening tool to rapidly identify phenotypic social extremes in a large population (n=5,200 monkeys), allowing these monkeys to serve as a model for early-life interventions, whether pharmacologic or behavioral. To assess the construct validity of this model, we will test whether in low-sociable vs. high-sociable male monkeys there is an excess of rare pathogenic variants in genes previously implicated in autism. We will also test whether candidate biomarkers of social functioning (e.g., oxytocin, arginine-vasopressin, extracellular regulated kinase and related kinases), as measured in the CSF and blood from these monkeys, are associated with sociability status, whether the degree of biomarker dysregulation co-varies with the degree of social deficits, and whether CSF and blood measures are equally informative. These findings will link abnormal social behavior with protein and gene based biological changes in a way that has not been previously achievable. We are optimistic that further development of this model will accelerate the discovery of autism biomarkers and novel drugable targets, provide new blood-based metrics for treatment response efficacy, and streamline the development of effective therapeutics that will benefit patients with presently intractable social impairments.